can the model handle any size test images?

[generallc]

Hi, thank you for your excellent research , your paper uses the 224 x 244 image for training, I would like to ask can the training model handle any size image?

[richzhang]

Yes, it can be fun fully convolutionally.

[generallc]

Thank you for your reply, I try to use keras to achieve, but still only test the fixed size，which is the same as the training input size, because I am not familiar with caffe, very curious about your caffe training model can test any size image? And I see you write a colorize.py, using resampling operations on different sizes of test images, why not directly enter the network?